Phosphorylation of SAMHD1 Thr592 increases C-terminal domain dynamics, tetramer dissociation and ssDNA binding kinetics

Abstract

SAM and HD domain containing deoxynucleoside triphosphate triphosphohydrolase 1 (SAMHD1) is driven into its activated tetramer form by binding of GTP activator and dNTP activators/substrates. In addition, the inactive monomeric and dimeric forms of the enzyme bind to single-stranded (ss) nucleic acids. During DNA replication SAMHD1 can be phosphorylated by CDK1 and CDK2 at its C-terminal threonine 592 (pSAMHD1), localizing the enzyme to stalled replication forks (RFs) to promote their restart. Although phosphorylation has only a small effect on the dNTPase activity and ssDNA binding affinity of SAMHD1, perturbation of the native T592 by phosphorylation decreased the thermal stability of tetrameric SAMHD1 and accelerated tetramer dissociation in the absence and presence of ssDNA (∼15-fold). In addition, we found that ssDNA binds competitively with GTP to the A1 site. A full-length SAMHD1 cryo-EM structure revealed substantial dynamics in the C-terminal domain (which contains T592), which could be modulated by phosphorylation. We propose that T592 phosphorylation increases tetramer dynamics and allows invasion of ssDNA into the A1 site and the previously characterized DNA binding surface at the dimer-dimer interface. These features are consistent with rapid and regiospecific inactivation of pSAMHD1 dNTPase at RFs or other sites of free ssDNA in cells.

Figure 1.

Both the T592E phosphomimetic mutation and phosphorylation have no significant impact on the dNTPase activity or DNA-binding affinity of SAMHD1. (A) Kinetic analysis of wild-type SAMHD1. SAMHD1 (0.5 μM) was incubated with varying concentrations of GTP and [2-¹⁴C]-dTTP. Curves represent least-squares regression fit to the Michaelis-Menten equation. (B) Kinetic analysis of SAMHD1 T592E under the same conditions as in (A). Error bars in (A) and (B) indicate standard error of reaction rate determined by the linear regression fit of the linear phase of three replicate reactions performed for each condition. (C) Secondary replot of apparent K_m with respect to dTTP substrate concentration as a function of GTP activator concentration. Error bars indicate standard error of the K_m values determined in (A) and (B). (D) dNTPase activity of pSAMHD1 compared with wild-type SAMHD1 and SAMHD1 T592E under select conditions from the kinetic analysis in (A) and (B). Error bars indicate standard error of the slope determined by the linear regression fit of the linear phase of three replicate reactions under each condition. (E) Binding of SAMHD1 wild-type T592E, and pSAMHD1 to 5′FAM-labeled ssDNA57 (50 nM). (F) Binding of SAMHD1 wild-type, T592E, and pSAMHD1 to 5′FAM-labeled PsDNA5 (50 nM). Error bars in (E) and (F) indicate the standard error of mean determined by three replicate titrations.

Figure 2.

GTP Activator and single-stranded DNA compete for the same binding site on SAMHD1. (A) Binding of SAMHD1 wild-type, T592E, and pSAMHD1 to 5′ FAM-labeled ssDNA57 in the presence of no nucleotides, increasing concentrations of GTP, and a combination of GTP/dTTPαS (1 mM each). (B) Apparent K_d values from (A) plotted as a function of GTP concentration. Error bars indicate standard error of K_app^DNA from least-squares regression fit of the data in (A) to Equation (5). (C) Isolated SAMHD1 monomer from the GTP & dATP bound tetrameric SAMHD1 structure (PDB 6TXC). DNA-interacting residues identified via BrdU-crosslinking are highlighted in red and cationic residues involved in DNA binding are highlighted in blue. (D) Isolated SAMHD1 monomer from a PsDNA5 bound SAMHD1 structure (PDB: 6U6X). DNA-interacting residues identified via BrdU-crosslinking are highlighted in red and cationic residues involved in DNA binding are highlighted in blue.

Figure 3.

Single-stranded DNA displaces N-methyl anthraniloyl GTP (mant-GTP) from the A1 site. (A) Emission spectra of mant-GTP (0.5 μM) alone and bound to 10 μM SAMHD1. Emission scans were taken from 380 nm to 500 nm with an excitation wavelength of 335 nm. (B) Activation of dTTP (1 mM) hydrolysis by mant-GTP (0.5 mM) as compared to the same concentration of GTP (Figure 1A). Error bars indicate standard errors determined by linear regression fitting to the linear phase of three replicate reactions. (C) Binding of mant-GTP (0.5 μM) to wild-type SAMHD1 in the presence of dATP (1 mM) or GTP (1 mM). (D) Binding of mant-GTP (0.5 μM) to wild-type SAMHD1 in the presence of ssDNA57 (20 μM) or psDNA5 (20 μM). The nonspecific binding measured in (C) was subtracted from each of the binding isotherms. (E) Binding of SAMHD1 T592E to mant-GTP (0.5 μM) in the presence and absence of ssDNA57 (20 μM) or psDNA5 (20 μM). The background nonspecific binding contribution was subtracted from each of the binding isotherms (Supplemental Figure S8A). (F) Binding of pSAMHD1 to mant-GTP (0.5 μM) in the presence and absence of ssDNA57 (20 μM) or psDNA5 (20 μM). The background nonspecific binding contribution was subtracted from each of the binding isotherms (Supplemental Figure S8B). Error bars in (C) through (F) represent standard errors of mean from three replicate titrations.

Figure 4.

Conformational flexibility of the C-terminal domain. (A) Plot of relative fluctuations by residue determined by coarse grained normal mode analysis on an isolated monomer from a tetrameric SAMHD1 structure (PDB: 6TXC) (black) and by alignment of isolated monomers from tetrameric structures of SAMHD1 T592V (PDB: 4ZWE) and SAMHD1 T592E (PDB: 4ZWG) (pink). Secondary structure is indicated at the top of the plot, with black bars indicating α-helices, and red bars indicating β-sheets. The C-terminal Domain (CtD) and flexible hinge region are labeled above the plot. (B) Fluctuations determined by normal mode analysis mapped on the structure of SAMHD1 (PDB: 6TXC). (C) Deformation energy determined by normal mode analysis mapped on the structure of SAMHD1 (PDB ID 6TXC). (D) Conformations of the C-terminal domain (CtD) (T592V – PDB: 4ZWE, T592E – PDB: 4ZWG). The T592E mutation causes the CtD to shift away from the allosteric sites and towards the active site. (E) Local resolution cryo-EM map of full-length hSAMHD1 bound to dGTPαS. The cryo-EM map is shown at a lower contour to highlight the lower resolution (blue) SAM and C-terminal domains. (F) An atomic model based on the SAMHD1 catalytic domain was docked into (E), showing the residual unmodeled densities corresponding to the SAM domains.

Figure 5.

Phosphorylation, phosphomimetic mutation, and elimination of the phosphorylation site decrease the thermal stability of the SAMHD1 tetramer. (A) Normalized first derivative plot of thermal melts of wild-type SAMHD1, T592E, pSAMHD1, SAMHD1 Δ583–626, and SAMHD1 Δ600–626. SAMHD1 (3 μM) was incubated with 2 mM dGTPαS to drive the system to a tetrameric state. Temperature was ramped up from 25°C to 85°C. (B) Melt temperatures of individual replicates from (A) calculated in TSA-craft. (C) Hydrogen bonding network formed by threonine 592, aspartate 585, and lysine 580 (PDB: 6TXC).

Figure 6.

Both phosphorylation and phosphomimetic mutation sensitize SAMHD1 to inhibition by single-stranded DNA. (A) Schematic of dilution experiment in (B), (C) and (D). A 25 μM SAMHD1 solution was preincubated for 30 s with 1 mM dGTP or with no dGTP, then diluted 100-fold into a solution of 50 nM 5′FAM labeled ssDNA57. Anisotropy of the FAM fluorophore was measured at 15 second intervals for 10 minutes. (B) ssDNA binding kinetics of SAMHD1 WT. (C) ssDNA binding kinetics of SAMHD1 T592E. (D) ssDNA binding kinetics of pSAMHD1. Error bars for (B), (C) and (D) indicate standard error of mean from 5 replicate reactions. (E) Inhibition of SAMHD1 dNTPase activity by increasing concentrations of single-stranded DNA. SAMHD1 (0.5 μM) was incubated with 10 μM dTTP, 10 μM GTP and 0–50 μM ssDNA90. Error bars represent standard error of the reaction rate as determined by linear regression analysis of the linear phase of three replicate reactions.

Figure 7.

SAMHD1 tetramer dissociation is enhanced by phosphomimetic mutation, phosphorylation, elimination of the phosphorylation site, and by the presence of single-stranded DNA. (A) Schematic of dilution experiment. SAMHD1 (25 μM) was incubated with 1 mM dGTP for 30 seconds, then rapidly diluted 100-fold into a solution containing no activating nucleotides, as in (B), or into a solution of 5 μM ssDNA90, as in (C), and crosslinked in glutaraldehyde (50 mM) at regular timer intervals to monitor dissociation of the tetrameric state. (B) Dissociation kinetics of wild-type SAMHD1, T592E, pSAMHD1, SAMHD1 Δ583–626 and SAMHD1 Δ600–626. Select excerpts of gels are shown on the left, and quantification of the three replicates done for each enzyme are shown on the right. Error bars represent standard error of mean from three replicate reactions. (C) Dissociation kinetics of wild-type SAMHD1, T592E, pSAMHD1 in the presence of 5 μM ssDNA90. Select excerpts of gels are shown on the left, and quantification of the three replicates done for each enzyme are shown on the right. Error bars indicate standard error of mean for three replicate reactions.

Figure 8.

A model for phosphorylation-dependent disassembly of SAMHD1 at stalled replication forks. Structural perturbation of the C-terminal domain (shown in purple) caused by phosphorylation decreases the stability of the tetrameric state. While this destabilization does not have a strong impact on the steady-state dNTPase activity of the enzyme in the presence of activating and substrate nucleotides, it does increase tetramer dynamics. The presence of ssDNA induces rapid dissociation of the pSAMHD1 tetramer into predominantly dimeric products. Due to the antagonistic nature of GTP and ssDNA binding, we presume that dissociation involves displacement of GTP by ssDNA strand invasion into the allosteric sites and DNA binding interface.